Mechanical interaction of intersecting normal faults: Theory, field examples and applications.

摘要

This dissertation investigates the mechanical interaction between coeval intersecting normal faults and the effect on fault slip characteristics, such as slip distribution and direction, by comparing observations of natural fault systems with three-dimensional numerical modeling. It then investigates two methodologies, using numerical modeling and three-dimensional seismic data, to refine fault continuity and linkage when the quality of seismic data is poor, and to predict faults below the seismic resolution in reservoir. The results of numerically-computed slip distributions from three-dimensional elastic models indicate that the asymmetric and irregular slip distributions and slip vectors observed on natural examples can be caused by mechanical interaction between intersecting faults that produce local perturbations of the stress field resolved on the faults. These results agree with observations on natural normal fault examples.; A method has been developed to infer, in three dimensions, the fault tip-line geometry below the seismic resolution, as well as potential fault linkage, using three-dimensional seismic data and geomechanical models based on elastic dislocations. By comparing the interpreted fault slip distribution to the computed slip distribution adjacent to potential intersection lines, the geomechanical models can constrain the geometry of the faults as well as the location of the intersection line between faults.; A second method also is proposed here for predicting the positions and orientations of subsesimic faults in the rock volume using mechanical modeling. The largest faults are brought into a three-dimensional numerical mechanical model in order to determine what the stress conditions were around the faults at the time of faulting. The stress field is then combined with the Coulomb failure criterion in order to predict the orientations and densities of the smaller faults; this information is represented as a pair of grids (e.g. density and strike grid). The grids are then used to condition a three-dimensional stochastic model of faulting, which uses a power-law distribution to determine the sizes of the faults.